Recommended Kit to obtain station coordinates using PPK

GPS/GNSS
1

One man band surveyor here in the UK. Doing measured building and topo (generally up to 2 acres). Running a Leica TS13 and BLK360.

Looking for a bargain basement way to georeference my jobs. Basically I need to obtain the positions of 2 station nails on each job in WGS84 / ETRS89 coordinates.

I’ve been getting away with arbitrary local datums on a lot of my jobs. Where possible I level to Ordnance Survey benchmarks. Occasionally a Flood Risk Assessment is needed and I’ve had to pay another surveyor £200 to come and do the GPS work. It’d be nice to be able to properly georeference all jobs myself.

I’ve looked at the Emlid Reach RS+ which is around £1000 here.
Also looked at the Sparkfun surveying kits which range from £450 to £1200. But these seem to be OEM/EMS only here?

Now looking at Sparkfun individual components.

From what I can tell, the Emlid Reach RS+ uses a Ublox M8U (I think), is Single Band, with 72 Channels. Would a Sparkfun board using a ZED-F9P actually be superior? From researching, the greater the number of channels should equate to being generally better at obtaining a good fix, especially in urban areas or with vegetation?

Am I correct, that at the most basic level, all I need is a RTK board, an antenna which I can mount on a tribrach/tripod, some connecting cables, and my phone/tablet? I can then log data for a period of time, and use PPK to process the data in the office to achieve cm accuracy?

Which board would be sufficient/recommended?
What app would I use to log the data?
How long should the data be logged for to achieve a decent result?
I could use RTKlib or Emlid Studio, to convert the logged data to RINEX and process using the relevant RINEX file downloaded from the Ordnance Survey website?

When getting a surveyor with GPS to come and obtain the coordinates, I’m generally getting a 5-10mm discrepancy for the Eastings/Northings, station to station as compared to my theodolite readings, and 30-40mm discrepancy in elevation. I usually just average the GPS readings and move the whole job to fit.

Any help/advice appreciated. Thanks.

2

Some of the specific answers to your questions will depend on the software that you use for scan registration.

I haven’t used my BLK360 very much after Leica pulled the “Rug Trick” with ReCAP 360, so my software knowledge will be a little dated.

General Thoughts:

While you can easily use any RTK Device (I’ve actually been thinking about the PostCard with my BLK360), I don’t think it will help “much” with the Scan-to-Scan Registration.

I’d consider performing PPP on 2 control points that you will also occupy with the BLK360, or at least occupy with targets.

Those 2 points can be used to manually perform the final 3D Model adjustment to situate it in whatever coordinate system you want. Your Scan Registration Software might allow you to set the Coordinates of your 2 Control Points to confine the 3D Model during project registration.

With an appropriate PPP workflow, you should be well under the 30-40mm in elevation discrepancy you mentioned.

That’s likely the Grid-to-Ground relationship that we all experience when comparing GNSS vs Conventional Surveying. Applying the Scale Factor for the project area is the solution when comparing Grid to Ground coordinates.

I’d consider (2) PostCards, Portability Shields, Li-Po’s, etc. Or simply (2) Facets if you don’t want to fool with building anything. Let both GNSS units log RAW data the entire time you are onsite. Pull the micro-SD Cards at the end of the mission and use CSRS-PPP (or your preference) to provide the (2) PPP solutions.

Which board would be sufficient/recommended? Facets are the easiest, PostCards are the cheapest and add L5, you can use (2) for concurrent logging of your (2) CP’s
What app would I use to log the data? None for PPP, RTCM is stored on SD card and you later convert to RINEX and submit for PPP.
How long should the data be logged for? Loaded Question ~6 hours for PPP, 24 hours isn’t uncommon- it depends on your goals and the environment.
I could use RTKlib or Emlid Studio, to convert the logged data to RINEX and process using the relevant RINEX file downloaded from the Ordnance Survey website? Yes

Side Note: when using GNSS for a Flood Risk Assessment, I’d want more than 1 PPP mission solution to verify the Delta from the BM to my local control. But the good news is you set it up and leave it.

Warning: Once you get a Survey-Grade GNSS and a quality RTK correction source, you will love using AutoTOPO and other various features. Your Leica Total Station will spend a lot more time in the truck :wink: You might decide that the Torch or a Facet Mosaic is a better solution for all your needs (in addition to Control Points for Laser Scanning).

There are several Professional Surveyors and others on this forum that will hopefully comment to give you some more opinions.

Oh, and WELCOME to the Community :slight_smile:

1 Like
3

Thanks for the response.

Haha, I briefly used Recap Pro (actually an Autodesk product and it’s still used by some, the mobile app was discontinued), hated it. I’ve been using Leica Register 360 since the plug was pulled, it’s far superior (if a tad expensive). You can create closed loops of registrations and it automatically distributes any errors. My registered clouds always fit to the control targets within mm throughout.

I always do a closed traverse with the theodolite and survey in multiple internal and external targets to control the point clouds. So I’ll usually have a few nails or stakes in the ground.

All my scanner setups are random positions. I was thinking only one device, and to occupy 2 points on the theodolite traverse for only an hour or less each (there are videos of surveyors doing this on Youtube).

Having 2 devices (double the cost) occupying stations for the whole day (no security and less flexibility) is not ideal… (I need to be able to setup the theodolite on these points as well).

So is PPP more accurate than PPK? I’m unfamiliar with the whole GNSS subject so a bit lost on the technicalities…

Yes, I’ve been longing for a Leica AP20 autopole (tilt compensation and auto pole height) but money is tight and work a bit intermittent (they’re around £7000), and I can see that a proper base/rover setup for topo is the way to go. It’s a money thing, and most of my jobs are quite small with the majority of each survey done using the scanner, so the TS13 is fine.

I’m into a bit of DIY so just a postcard is also fine. Really want to do this as cheap as possible, almost on a hobby basis, but still usable to orientate/level actual surveys.

4

Wow, thanks for that info. I always applied my control at the end and just used targets for Scan-to-Scan registration and as a QC check.

I concur, but we’re talking about ~$300 for a PostCard and Antenna which makes it a littler easier to leave unattended for a bit, verses a $10k GNSS device for the same performance.

Similar but different approaches. PPP is for a Static Position, let it cook for a long time to get the best result possible for 1 point in space. PPK could be described as an alternative to RTK when a real-time correction source wasn’t available during the mission (that’s not technically 100% accurate but a decent starting point for comparison)

For TOPO, any of the RTK products at Sparkfun (including the PostCard) will work fine.
The simple solution is have an online RTK correction Source over Cellular, say NTRIP for example .
Take Topo shots as quick as you can walk (or ride) :wink:
In most cases, I can get comfortable considering 1 second Topo shots to be <1cm accuracy.
That’s qualifying the correction source and overall performance each day.
The Correction Source and Antenna have more impact than the specific L1/L2/L5 GNSS chipset in this context, IMHO.

You’ve made me go charge up the batteries for the BLK360 :slight_smile:
Thanks !

1 Like
5

I love my BLK360, but it’s a G1, every time I use it I think of the G2 (bought mine 5 years ago before the G2 was released) which is approx 5 times faster, so you can be talking seconds per scan rather than minutes, akin to the RTC360.

6

Hello I’ll add to the engineer Ryan above as I am surveyor Ryan (btw I sit for the PE in 4 months Ryan). I have been playing with facet and postcard and I am still running a turn screw Nikon for conventional. I would say hands down for budget the postcard with some learning curve can do pretty much everything you need. I find myself wanting to buy a couple torches but I will be testing those soon through a demo with sparkfun.

In my day to day at work I run a Trimble s7 and r12i and sometimes a sx12 scanner. Like you I run traverses etc.

Where something like the postcard really makes sense for me is getting my control into coordinates so I can leverage GIS and other things for boundary surveying. I still use my Nikon to shoot pretty much everything to control accuracy. But for instance I’m going to go locate property corners today, I can just break my phone out with surpad, use calculated coordinates and get within usually 5 feet or less.

Today I am doing a job where we don’t have cell service or the luxury of our statewide RTN network so I might be trying using survey in base and LoRa radios.

The site is also a forest so the gps might not be useful at all today.

I guess what I am saying at approx 400 USD for a postcard setup you can’t really go wrong adding it as a tool.

I have case files Steven from sparkfun gave me if you want them.

20250209_063141
20250209_0631411920×2560 267 KB

2 Likes
7

Thanks for the input.

I think I’m gonna go for the postcard / portability shield combo and a good antenna.

Do you have your setup with a battery sandwiched between the two boards, as shown on the Sparkfun website?

Also, are the log files on the SD card downloaded via USB, not by removing the SD card and plugging it into a card reader?

Like you say, at the price, I can’t really go wrong.

The case files would be nice. I was thinking about what to do about a case.

Thanks.

8

I actually use a bigger battery for 30 hours of life and the case is designed for that. You can pull logs via webui or I just remove the card and use a card reader.

I will say I had a boundary job today that had trees etc and I think my facet honestly held out better being even an l1/l2 just because of gain? However I still got decent fixes with the postcard and got myself within about 3 feet of monuments from my calc. So what am I complaining about?

I did use my facet to throw a quick 500 foot baseline into projected coordinates and my total station checked between the two by 0.02 horizontally. Obviously with boundary I don’t care about vertical really. We also are lucky enough to have low distortion projections where grid basically equals ground. I still held conventional measurements but 2 3 minute observations got my baseline spun into the same reference frame as my projected coordinates rather than classic 10000,10000 etc.

Now granted we have this what I consider insanely good RTN where it takes 3 cors around you averages on the fly and spits the corrections to you so you don’t see as much error as using a single base.

I think I’m gonna order an facet mosiac after what I saw today. Shame it isn’t waterproof or doesn’t include an imu. But although I have 2 RTK postcards setups and lora I think I will get a facet mosiac for what I would assume would be a strong antenna gain.

1 Like
9

Thanks for the further insights. Is it possible to elaborate on “I will say I had a boundary job today that had trees etc and I think my facet honestly held out better being even an l1/l2 just because of gain?”

On paper (from my amateur research) the postcard (with the same antenna) should be superior?

Facet - u-blox F9 - 184 channels

Postcard - Quectel LG290P - 1040 channels

The postcard has a lot more channels, which from what I’ve read helps with getting a fix in challenging conditions.

“…housed under the dome of the RTK Facet is a surveyor grade L1/L2 antenna. It is the same element found within our GNSS Multi-Band L1/L2 Surveying Antenna.”

Facet - Antenna - L1/L2 with ≥5dBi gain. LNA Gain: 40dB
GNSS Multi-Band L1/L2 Surveying Antenna - TNC (TOP106)

Sparkfun latest antenna - GNSS Multi-Band L1/L2/L5 Surveying Antenna - TNC (SPK6618H)

Reception Frequencies 1150 to 1300MHz 1520 to 1610MHz Supported Constellation Signals GPS - L1/L2/L5 GLONASS - G1/G2 BeiDou - B1/B2/B3 Galileo - E1/E5a/E5b/E6 LNA Gain: 38dB Gain: ≥2.5dBi GPS L1 ≥5.5dBi L2 ≥5.0dBi L5 ≥2.5dBi GLONASS G1 ≥5.0dBi G2 ≥4.0dBi BeiDou B1 ≥5.0dBi B2 ≥5.0dBi B3 ≥3.5dBi Galileo E1 ≥5.5dBi E5a ≥2.5dBi E5b ≥5.0dBi E6 ≥2.5dBi 50 Ohm Antenna Impedance

Further research on LG290P vs ublox f9 -

“The LG290P is specifically designed with advanced algorithms to combat signal interference in complex environments like urban areas or dense foliage, potentially providing better accuracy in difficult conditions compared to the ublox F9. … If your application requires high accuracy in areas with significant signal interference, the LG290P might be a better choice due to its advanced interference mitigation features.”

It’s entirely possible that I’m just “doing my own research” and getting this all wrong. It’s actual results in the field that count.

10

@rftop has the opposite results (he too sees better with postcard). I own both but I’m gonna go give it some more tests. I was in heavy trees yesterday and it seemed like my facet held better. One advantage having that big red facet on top of the prism pole and doing integrated surveying basically was I could see the darn prism through the trees yesterday.

1 Like
11

I hate doing trees with my TS13. Only been on my own for 5 years with the total station. Previous to that I worked for another surveyor. We had a reflectorless Leica instrument but for topo work it was an old Geodometer from the 80s and a Psion Organiser for logging. I was on the instrument with the logger and the other guy would wander about in the trees with the prism pole shouting out offsets and tree info. It was amazing how the Geodometer could see through the leaves and get an instant reading on track mode (I think it was infra red).

12

I’m running a Nikon dtm 322, I want to get some kind of leica for a robot but after yesterday I cannot ever fathom getting rid of the Nikon.